Figure A contains a plot of 200 pieces of data from ten card subsets of a single deck and provides insight into the behavior of least square estimates. Note the boid, parabolic, nature suggested for the regression function, similar to the shape wei would observe if i plotted the 13 card regression function for the Woolworth games.[B]

Other data gathered in these experiments shed light on the `Count of Zero' phenomenon. The player's actual advantage, as a function of two ill known card counting systems, the Ten Count and the Hi Lo, also displayed the same parabolic shape. Consequently counts near zero, reflecting normal proportions of unplayed cards, had actual expectations higher than linear theory would predict. For example, with 13 cards remaining, a Hi Lo count of zero was associated with an expectation of +1.60%, while 13 card subsets with precisely four tens remaining had a player advantage of +1.87%.

The most probable 13 card subset, one card of each denomination, had a 2.05% expectation for single deck basic strategy, 2.07% above the linear estimate of -.02%. The

highest estimated expectation, 27.36%, occurs for 4 aces and 9 tens and is 3.44% above the actual expectation, while 4 each fours, fives, and sixes, and one three has the most negative estimated expectation of -28.39%, 138.03% above the actual value.